TL;DR: ⚪ Null — liberal transfusion did not clearly improve neurologic outcomes specifically in aneurysmal SAH (33.5% vs 37.7%, P=0.22) — contrasts interestingly with TRAIN's positive result in broader brain injury.
1. Publication
- Title: Liberal or Restrictive Transfusion Strategy in Aneurysmal Subarachnoid Hemorrhage
- Acronym: SAHaRA
- Year & Journal: New England Journal of Medicine, published December 9, 2024 (2025;392(11):1079-1088)
- Citation: English SW, Delaney A, Fergusson DA, et al; SAHaRA Trial Investigators. N Engl J Med. 2025;392(11):1079-1088. doi:10.1056/NEJMoa2410962
2. Context & Rationale
Background: Anemia occurs in ~50% of aneurysmal SAH patients and is linked to worse outcomes, plausibly via reduced brain tissue oxygen delivery. Evidence on optimal transfusion threshold in this population is limited and conflicting, leading to vague guideline recommendations.
Research Question/Hypothesis: In critically ill adults with aneurysmal SAH and anemia, does a liberal transfusion strategy reduce unfavorable neurologic outcome at 12 months compared with a restrictive strategy?
3. Design & Methods
- Study Type: Multicenter, open-label RCT with blinded outcome assessment
- Setting & Centers: 23 centers, Australia/Canada/US
- Population: 725 analyzed adults (of 742 enrolled) with acute aneurysmal SAH and anemia; mean age 59.4y, 82% women
- Intervention: Liberal strategy — mandatory transfusion at Hb ≤10 g/dL (n=364)
- Comparator: Restrictive strategy — optional transfusion at Hb ≤8 g/dL (n=361)
- Blinding: Open-label treatment, blinded outcome assessment
- Statistical Power & Follow-Up: Primary: unfavorable neurologic outcome (mRS≥4) at 12 months.
4. Key Results
725 patients analyzed.
Outcome | Liberal | Restrictive | Effect Size | p-value | Notes |
Unfavorable neurologic outcome at 12mo (primary) | 33.5% (122/364) | 37.7% (136/361) | RR 0.88 (95% CI 0.72–1.09) | 0.22 | No significant difference |
5. Internal Validity Assessment
Large (725-patient), multicenter RCT with blinded outcome assessment despite open-label treatment. Overall: Strong — clean, precise null result with good follow-up completeness.
6. External Validity Assessment
23-center, 3-country (Australia, Canada, US) SAH population with neurosurgical/neurointerventional support — broadly representative of aSAH populations managed in high-resource centers.
7. Strengths & Limitations
Strengths: Large, multicenter, blinded outcome assessment; addresses a genuine, guideline-relevant uncertainty.
Limitations: Open-label treatment delivery; authors note value in integrating these results with other research (e.g., TRAIN, this handbook) for a fuller picture across brain injury subtypes.
8. Interpretation & Practice Impact
Does not support liberal transfusion over restrictive transfusion in aSAH — contrasts notably with TRAIN's finding (this handbook) of benefit from liberal transfusion in a broader acute brain injury population (TBI, SAH, ICH combined), an important nuance for interpreting transfusion strategy across different brain injury subtypes.
9. Controversies & Subsequent Evidence
A subsequent systematic review/meta-analysis ("Approaches to Guide Red Blood Cell Transfusion in Patients with Subarachnoid Hemorrhage," J Intensive Care Med 2026) synthesizes this and related evidence. The apparent discordance between SAHaRA (null, SAH-specific) and TRAIN (positive for liberal, broader acute brain injury population including SAH) is a genuinely important area for further subgroup-level investigation.
10. Summary & Executive Takeaway
Summary: SAHaRA randomized 725 aneurysmal SAH patients with anemia across 23 centers to liberal (Hb≤10) or restrictive (Hb≤8) transfusion. Unfavorable neurologic outcome at 12 months was similar (33.5% vs 37.7%, RR 0.88, P=0.22).
Overall Takeaway: Liberal transfusion does not clearly improve neurologic outcomes specifically in aneurysmal SAH — an important, SAH-specific null result that stands in interesting contrast to TRAIN's positive finding for liberal transfusion in the broader acute brain injury population, suggesting transfusion strategy effects may vary by brain injury subtype.
11. Bibliography
- Taccone FS, et al. TRAIN Randomized Clinical Trial. JAMA. 2024;332(19):1623-1633 (this handbook).
- Hebert PC, et al. TRICC trial. N Engl J Med. 1999;340:409-417.